Anal. Chem. 2000, 72, 4381-4385
Chemical Grafting of Molecularly Imprinted Homopolymers to the Surface of Microplates. Application of Artificial Adrenergic Receptor in Enzyme-Linked Assay for β-Agonists Determination Sergey A. Piletsky,* Elena V. Piletska, Beining Chen, Khalku Karim, David Weston, Gary Barrett, Phillip Lowe, and Anthony P. F. Turner
Institute of BioScience and Technology, Cranfield University, Bedfordshire, MK43 0AL, UK
A technique for coating of microplate wells with a molecularly imprinted polymer (MIP), specific for epinephrine, is presented. 3-Aminophenylboronic acid was polymerized in the presence of epinephrine using oxidation of the monomer by ammonium persulfate. This process resulted in the grafting of a thin polymer layer onto the polystyrene surface of the microplates. The polymer affinity was determined by an enzyme-linked assay using a conjugate of horseradish peroxidase and norepinephrine (HRP-N). It was found that imprinting resulted in increased affinity of the polymer toward HRP-N and epinephrine. Influence of the buffer pH and concentration on the polymer affinity was analyzed. It was shown that the MIP-coated microplates could be used for assay development and drug screening. The high stability of the polymers and good reproducibility of the measurements make MIP coating an attractive alternative to traditional antibodies or receptors, used in ELISA. Bioassays, in particular immunoassays or receptor-based assays, are routinely used in clinical, environmental, agricultural/ food, and forensic industries for the analysis of proteins, hormones, viruses and microorganisms, DNA sequences, and drugs.1,2 The enzyme-linked immunoassay (ELISA) is probably the most common method, which utilizes the competition between free and enzyme-labeled ligands to immobilized antibodies for the analyte monitoring.3 Immunoassays are rapid, sensitive and selective to the analyte of interest, and generally cost-effective for large sample loads. However, as with any technology, there are disadvantages; for example, the reagent stability and the high expense of producing antibodies are often cited as problems. In this regard, molecularly imprinted polymers have already been identified as stable receptor or enzyme mimics, suitable for substitution for the natural receptors in assays or sensors.4-6 * To whom the correspondence should be addressed. Tel: + 44(0)1234 754339 ext.3584. Fax: +44(0)1234 750907. E-mail:
[email protected]; http://www.cranfield.ac.uk/ibst. (1) Tremblay, L.; VanderKraak, G. Aquat. Toxicol. 1998, 43, 149-162. (2) Nakanishi, K.; Huang, X. F.; Jiang, H.; Liu, Y.; Fang, K.; Huang, D. W.; Choi, S. K.; Katz, E.; Eldefrawi, M. Bioorg. Med. Chem. 1997, 5, 1969-1988. (3) Viera, A. Mol. Biotechnol. 1998, 10, 247-250. 10.1021/ac0002184 CCC: $19.00 Published on Web 08/10/2000
© 2000 American Chemical Society
Their inherent stability, low cost, and ease of preparation7,8 offer several major advantages over the use of natural receptors and antibodies. However, MIPs have shortcomings, such as a lack of true water compatibility with most polymerization systems thus far reported as well as difficult immobilization procedures. In particular, the absence of a reproducible method of coating microplate wells with MIPs impairs their application in assays where this format is preferable. It is possible to prepare a uniform coating of virtually any support with a thin layer of conjugated polymer using chemical grafting or electropolymerization.9-11 Additionally, conjugated polymers were described as promising materials, suited for MIP formation.12 Here we present the combination of these two approaches in a new method of microplate modification with imprinted homopolymer- poly-3-aminophenylboronic acid. Epinephrine, a natural ligand of the adrenergic receptor, was selected as a template for these experiments. This work is part of research initiated in our laboratory and aimed at the synthesis and investigation of conjugated MIPs (both homo- and heteropolymers). Besides theoretical interest, this work has practical significance aimed at the development of a system for drug screening and assay. MATERIALS AND METHODS 3-Aminophenylboronic acid monohydrate (APBA), ammonium persulfate ((NH4)2S2O8), glutaric dialdehyde, (R)-(+)-propranolol hydrochloride, and (S)-(-)-propranolol hydrochloride were purchased from Aldrich. 2,2′-Azino-bis(3-ethyl)benzthiazoline-6-sulfonic acid (ABTS), catechol, clenbuterol hydrochloride, (-)(4) Andersson, L. I.; Mu ¨ ller, R.; Vlatakis, G.; Mosbach, K. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 4788-4792. (5) Vlatakis, G.; Andersson, L. I.; Mu ¨ ller, R.; Mosbach, K. Nature (London) 1993, 361, 645-647. (6) Piletsky, S. A.; Piletskaya, E. V.; Panasyuk, T. L.; El’skaya, A. V.; Levi, R.; Karube, I.; Wulff G. Macromolecules 1998, 31, 2137-2140. (7) Kriz, D.; Mosbach, K. Anal. Chim. Acta 1994, 300, 71-75. (8) Wulff, G. Angew. Chem., Int. Ed. Engl. 1995, 34, 1812-1832. (9) Kinlen, P. J.; Solverman, D. C.; Jeffreys, C. R. Synth. Met. 1997, 85, 13271332. (10) Pringsheim, E.; Terpetschnig, E.; Wolfbeis, O. S. Anal. Chim. Acta 1997, 357, 247-252. (11) Piletsky, S. A.; Kurys, Y. I.; Rachkov, A. E.; El’skaya, A. V. Russ. J. Electrochem. 1994, 30, 990-992. (12) Boyle, A.; Genies, E. M.; Lapkowski, M. Synth. Met. 1989, 28, C769-C774.
Analytical Chemistry, Vol. 72, No. 18, September 15, 2000 4381
Figure 1. Possible structure of the epinephrine-poly-APBA complex and structure of the analytes used in the assay.
epinephrine, horseradish peroxidase (EC1.11.1.7 from horseradish roots, MW 44 000 Da), (-)-isoproterenol (+)-bitartrate salt, (+)isoproterenol (+)-bitartrate salt, L-phenylephrine hydrochloride, pindolol, and norepinephrine (+)-bitartrate salt were purchased from Sigma. Polystyrene cuvettes (10 × 10 × 45 mm3) were purchased from Sarstedt (Numbrecht, Germany). Polystyrene microplates were purchased from Corning (Corning, NY). Modification of the Microplates with APBA/Epinephrine. A 50-µL solution of APBA (100 mM), containing 10 mM epinephrine, was placed in microplate wells using a multichannel pipet and mixed at room temperature with 50 µL of solution of (NH4)2S2O8 (50 mM). The polymerization reaction was carried out with gentle stirring for 5-60 min. After polymerization, the microplates were then washed thoroughly with 10 mM HCl solution and deionized water. Brown transparent films were formed on the wells. Blank polymer was formed under the same conditions in the absence of the template. HRP-N Conjugate Preparation. A sample of 4 mg of horseradish peroxidase was dissolved in 1 mL of norepinephrine solution in water (1 mM). Five microliters of glutaric dialdehyde solution in water (10 wt. %) was added to the mixture and incubated for 1 h at room temperature. The reaction was stopped by addition of 200 µL of lysine in water (50 mg/mL). The solution was filtered through an Ultra-Spin Macrofilter (Roth, Germany), separating molecules with molecular weight > 10 kDa, and 4382
Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
repeatedly washed on the filter three times with Na phosphate buffer (25 mM, pH 7.5). The activity of the conjugated enzyme was found to be 440 units per milligram of solid (determined using H2O2/ABTS). Analysis of the HRP-N Binding to the Polymer-Coated Microplates. The polymer-binding properties were analyzed using adsorption of horseradish peroxidase conjugate with norepinephrine (HRP-N) on the microplate well surface. This was done with seven concentrations ranging from 10 to 750 ng/mL of conjugate. After incubation for 1 h at room temperature, the microplate was washed three times with 0.005 wt. % Tween-20 and Na phosphate buffer (25 mM, pH 7.5). The reaction of HRP with hydrogen peroxide and ABTS was used to determine the concentration of conjugate adsorbed on the polymer. Six milligrams of ABTS solution was dissolved in 10 mL of Na citrate buffer (0.1 M, pH 6.0) and mixed with 3 µL of 30 wt % H2O2. One hundred microliters of solution was added to each well and incubated for 35 min at room temperature. Measurements were repeated every 5 min at 450 nm using a MR700 Microplate Reader (Dinex Technologies Inc., Chantilly, VA). All analyses were made at least in triplicate. Assay. Binding of the β-receptor agonists to the microplates, coated by polymer, during 20 min was measured using a competitive reaction between HRP-N conjugate and free analyte. A 50-µL aliquot of corresponding analyte (0-1 mM concentrations) and 50 µL of the HRP-N conjugate (625 ng/mL) were mixed
in the microplate wells and incubated for 1 h at room temperature. Washing and reaction with H2O2/ABTS were performed as described above. RESULTS AND DISCUSSION Formation of the stable complex between functional monomers and the template is crucial for the development of specific molecularly imprinted polymers. In our experiments, epinephrine was used as a template. This molecule possesses several functional groups, which are binding points for the interaction with the functional monomer: catechol group in the aromatic ring, hydroxyl group, and secondary amino group (Figure 1). This structure predetermined the choice of functional monomer, which in our case was 3-aminophenylboronic acid. Depending on solvent polarity and pH, poly-3-aminophenylboronic acid is able to form reversible covalent bonds between boronic acid and the catechol group in the template, ionic interaction between boronic acid and the secondary amino group, and several hydrogen bonds (Figure 1). Within the pH range 4.0-7.0, epinephrine is positively charged. As the pK′ of the boronic acid is 4.7 (determined by potentiometric titration of the 3-aminophenylboronic acid), it can be expected that the nature of interactions between the polymer and template changes from predominantly hydrophobic interactions and hydrogen bonds at pH < 5.0 to the combination of electrostatic, hydrophobic, and reversible covalent interactions at pH 8.0. The 1H NMR spectroscopic data indicated that the boronic acid and epinephrine indeed form a boronate ester at neutral pHs, although a significant part of the molecule was free and able to react with itself through electrostatic interactions. The MIP deposition on the microplate well surface was performed using a chemical oxidation process initiated by ammonium persulfate. The pH of the solution was adjusted to pH 4.0-6.0 using 10 M hydrochloric acid and to pH 7.0 and 8.0 using 2 M sodium hydroxide. The polymer deposition was time-dependent, reaching a maximum polymer thickness after 90 min (Figure 2a). The structure of the imprinted and blank polymers deposited on the microplate surface was visualized using AFM (data not shown). Both polymers gave almost identical relatively smooth reproducible surface coating with an average surface height variance of 40 nm. Grafting can therefore be seen to have little physical effect on the surface morphology. The polymer deposition was also dependent on the pH of the solution, with maximum deposition at pH 5.0 and minimum at pH 7.0 (Figure 2b). The polymer coating was transparent and homogeneous. The deviation in the polymer optical density between the wells of the microplate was less than 5%. The washing procedure is very important for assay development, and it was analyzed in experiments where buffers with different pHs (in the presence and absence of detergents) and different times of the treatment were applied. The optimized
Figure 2. Influence of the polymerization conditions: (a) time and (b) pH on the poly-APBA grafting to the polystyrene surface.
washing conditions, judging from the affinity of washed polymers, were found to be 5 cycles of washing with 10 mM HCl solution and, following that, 5 cycles of washing with deionized water. The polymer-binding properties were analyzed using adsorption of the horseradish peroxidase conjugate with norepinephrine (HRP-N) on the microplate well surface. Binding of the HRP-N conjugate to the MIP-coated microplate was dependent on buffer pH and concentration. The strongest response was obtained at pH 5.0, which gradually decreased with increasing pH. The dissociation constant was evaluated using Scatchard analysis. The resulting data clearly indicate the pH dependence of polymer specificity (Table 1). With an increase in the pH of the incubation buffer, binding of HRP-N conjugate decreased, reaching a minimum at pH 7.0. At pH 8.0, HRP-N conjugate binding increased, again as a result of reversible covalent bond formation between the polymer and template, which are more stable in basic solution. Microplates modified by MIP at pH 6.8 and 8.0 possess slightly improved binding characteristics in comparison with microplates modified
Table 1. Dissociation Constants of the HRP-N and HRP and Imprinted/Blank Polymersa
a
pH
5.0
6.0
6.8
8.0
MIP/HRP-N blank/HRP-N MIP/HRP blank/HRP
19.2 ( 0.6 nM 316 ( 11 nM 78 ( 2 nM 590 ( 20 nM
41.4 ( 2.4 nM 114 ( 27 nM 329 ( 7 nM 610 ( 10 nM
83.4 ( 2.8 nM 227 ( 10 nM 884 ( 18 nM 1250 ( 70 nM
65.0 ( 3.5 nM 124 ( 4 nM 1890 ( 20 nM 2150 ( 65 nM
Binding was analyzed using 50 mM Na phosphate buffers. Values given are means ( standard error measurements, obtained in triplicate.
Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
4383
at pH 4.0 and 5.0. It is possible to conclude that, in order to maintain high specificity, the polymer should be made under the same conditions as those applied during the assay. To evaluate the degree to which the conjugate binding is the result of interaction between the norepinephrine part of the molecule and specific binding sites, an adsorption analysis was performed using HRP-N and microplates, modified with blank polymer. Additionally, adsorption of enzyme itself was analyzed on the imprinted and blank polymers. The result of these measurements are presented in Table 1. At pH 5.0, grafted polymer is positively charged and efficiently binds negatively charged peroxidase. This charge is compensated by boronic acid, ionized at pH 6.0-7.0, leading to a decrease in peroxidase adsorption. At pH 8.0, the polymer becomes negatively charged due to secondary amino group deionization and is therefore no longer able to bind negatively charged peroxidase molecules. It is possible to conclude that enzyme-linked assay, based on polyAPBA, should be performed at pH 6.0-8.0 in order to minimize the nonspecific interaction between the enzyme and polymer. A significant difference between HRP-N and HRP binding to the polymers in favor of the conjugate clearly suggests that the norepinephrine part of molecule plays an essential role in the binding. In all cases, imprinted polymers had much higher affinity for the HRP-N conjugate than blank polymers, which is a good indication of the imprinting process and its influence on the formation of the template-specific binding sites. MIP was also able to bind HRP slightly better than blank polymer but this effect was much less pronounced and could be explained by higher porosity usually observed for imprinted materials. It was assumed that the structure of the specific binding sites in the MIP is preserved because of two factors: partial crosslinking of the polymers during polymerization and fixation of the polymer chains to the microplate well surface. Detailed analysis of the role of these factors in MIP specificity is under investigation. Microplates, coated by epinephrine-imprinted polymer, were tested in an enzyme-linked assay using competition between free ligand and HRP-N conjugate. Better results were obtained for the microplates coated with polymer for 20 min, due to minimal polymer interference in the color development. Taking into account the relative strength of the specific and nonspecific interactions between polymer and HRP-N conjugate and the magnitude of the sensor response, all following measurements were made in 50 mM Na phosphate buffer (pH 6.0). The result, presented in Figure 3 clearly indicates high specificity of the polymer for epinephrine. Ligands with only slightly different structures from the epinephrine had much weaker binding to the MIP than the template (Figure 3 and Table 2).
Figure 3. Displacement of the HRP-N by epinephrine (1), phenylephrine (2), (-)isoproterenol (3), (+) isoproterenol (4), norepinephrine (5), and catechol (6) from EPN-imprinted polymer. The microplate was tested in 50 mM Na phosphate buffer, pH 6.0.
The assay sensitivity is in the range 1-100 µM, which can probably be increased with further optimization of the polymer composition and assay performance. The dissociation constant for MIP-epinephrine is 9.2 ( 2 µM. This value is in the range of data found for natural adenilate cyclase coupled β-adrenergic receptor.13 Lower binding of epinephrine in comparison with HRP-N probably resulted from “unfair” competition due to multipoint binding of the conjugate to the polymer. When blank polymer was used instead of MIP, the dissociation constant decreased to 150 ( 15 µM (Table 2). The difference in epinephrine binding to the MIP and to the blank polymer clearly indicates the role of the imprinting process in the formation of specific binding sites for the template. It is important that if another template, such as phenylephrine, was used for polymerization instead of epinephrine, the specificity is inverted (for epinephrine KD ) 46 ( 5 µM and for phenylephrine KD ) 25 ( 2 µM) (Table 2). Measurements made with antagonists revealed much higher KD values than those found for agonists (for propranolol KD ) 157 ( 4 µM, for clenbuterol KD ) 300 ( 15 µM, and for pindolol KD ) 464 ( 33 µM). This is not a surprising result since antagonists do not have catechol or phenolic functionality. It is interesting, however, that the antagonists had similar affinity for both MIP and blank polymers. It can be concluded that MIP is able to discriminate between agonists and antagonists, which has not been achieved previously in direct-binding studies using natural receptors.13,14 The testing with samples of blood serum containing spiked concentrations of epinephrine demonstrated a significant decrease
Table 2. Dissociation Constants for MIP, Blank Polymers, and β-Receptor Agonists, Calculated Using a Competitive Assay in 50 mM Na Phosphate Buffer, pH 6.0a material
(-)epinephrine
(-)isoproterenol
(-)norepinephrine
(-)phenylephrine
MIP/Epinephrine MIP/Phenylephrine Blank Natural receptor13 Natural receptor14
9.2 ( 2 µM 46 ( 5 µM 150 ( 15 µM 4.6 ( 0.2 µM 50 nM
21 ( 2 µM 35 ( 2 µM 27 ( 1 µM 0.4 ( 0.01 µM 2.5 nM
46( 4 µM 107 ( 11 µM 90 ( 6 µM 49 ( 2 µM 2500 nM
>1000 µM 25 ( 2 µM 100 ( 5 µM 52 ( 4 µM
a The apparent K values were calculated from the equation K ) IC /(1 + [conjugate]/K conjugate). The K conjugate value was 41.4 nM. D D 50 D D Values given are means ( standard error measurements, obtained in triplicate.
4384 Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
in assay sensitivity. Obviously the presence of high amounts of sugars and proteins is responsible for the blocking of binding sites which limits the area of assay application to drug screening and testing of nonbiological samples. Further experiments were performed in order to evaluate the polymer stability. Binding of the HRP-N conjugate to the MIP has been evaluated in repetitive measurements carried out over two months. It was found that over this period of time the polymer retained 81% of its original activity and could still be used in an assay. CONCLUSIONS This study demonstrates that microplates can be coated with an imprinted polymer specific for epinephrine using polymerization of 3-aminophenylboronic acid with ammonium persulfate in (13) Mukherjee, C.; Caron, M. G.; Mullikin, D.; Lefkowitz, R. J. Mol. Pharmacol. 1975, 12, 16-31. (14) Harden, T. K.; Wolfe, B. B.; Molinoff, P. B. Mol. Pharmacol. 1975, 12, 1-15.
the presence of a template. The process results in the grafting of a thin polymer layer to the polystyrene surface of the microplates. Using an enzyme-linked assay, it was found that imprinting resulted in an increase of the polymer affinity toward HRP-N and epinephrine. Both buffer pH and concentration affect the polymer affinity, modulating the strength of the electrostatic and reversible covalent interactions. It is anticipated that MIP-coated microplates could be particularly useful for the development of assays and drug screening. ACKNOWLEDGMENT S.P. acknowledges with gratitude the fellowship from Leverhulme Trust.
Received for review February 18, 2000. Accepted June 22, 2000. AC0002184
Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
4385